8078 (9) Several anionic bases are known to catalyze H-D exchange of hydrogen: E. A. Symons and E. Buncel, Can. J. Chem, 51, 1673 (1973),and references therein. Surface phenomena may play a role in generating a microenvironment suitable for catalytic exchange which might not be duplicable in bulk solution. Also, the possibility of impurities should not be overlooked. (10) National Science Foundation Postdoctoral Fellow. Address correspondence to this author, Department of Chemistry, Harvard University, Cambridge, Mass. 02138.
Richard William Johnson,*lo Eric R. Holm Department of Chemistry, Stanford University Stanford, California 94305 Receiljed August 3, 1977
mm)), A,, (1:l .HCl-C2HsOH) 267 nm. Of the reagents previously used for the oxidation of thiazoline~,~ only activated M n 0 2 (CHC13, room temperature, 4 days) gave significant conversion of 3a to 4a; the latter was obtained as colorless crystals in 65% yield. A much better yield of 4a (93%) was obtained by the use of NiO2. In a typical experiment 293 mg (1.20 mmol) of 3a and 762 mg of Ni02* in 25 mL of CHC13 was shaken for 42 h. After filtration, concentration of the filtrate and crystallization of the residue (ether) gave 4a in a good state of purityg as colorless needles: mp 83-84 “C; Amax (C2HsOH) 236 nm; N M R (CDC13, (CH3)dSi) 6 1.45 (t, 3), 2.00 (s, 3), 3.28 (t, 2), 3.74 (m, 2), 4.42 (q, 2) 6.70 (br, I), 8.09 (s, 1). Analogous conversion of 2b to 4b was also effected, although the transformation 2b (R’ = H ) 3b generally proceeded in somewhat lower yield than 2a 3a. Saponification of 4a and 4b (KOH, aqueous dioxane) gave the respective carboxylates in yields of 96 and 95%. While the carboxylate derived from 4a had appreciable solubility only in water, and could not be condensed conveniently with S tritylcysteine ethyl ester, condensation of the acid derived from 4b with S-tritylcysteine ethyl ester (N,N’-dicyclohexylcarbodiimide, tetrahydrofuran) afforded tripeptide analogue 5b
--
A Biomimetic Synthesis of the Bithiazole Moiety of Bleomycin Sir.
The antibiotic bleomycin is of current interest because of its clinically useful anticancer activity.’ As part of a total synthesis of bleomycin B2 (l),we have been investigating the
5
6
RS’ 7
OANH~ 1
chemistry of the bithiazole moiety, the biosynthetic elaboration of which probably involves dehydrative cyclization of @-alanylcysteinylcysteine and dehydrogenation of the intermediate A2-thiazolines.2 Although the preparation of A2-thiazolines from certain cysteinyl peptides has been reported not to be po~sible,~ and no efficient methods have been recorded for the oxidation of complex A2-thiazolines, we report herein a biomimetic synthesis of the bithiazole moiety of bleomycin. Since several other natural products contain single thiazoles or A2-thiazoline group^,^ this synthetic approach should also be of more general utility. Although several agents previously employed for the preparation of simple thiazolines5 failed to effect the conversion of dipeptide 2a6 to the corresponding thiazoline, treatment of
0 a , R’CH3
b,R=CgHg
“ N u - : \
4
chloroform solutions of 2a (R’ = H or (C6H5)3C) with hydrogen chloride at 0 OC afforded ethyl 2-(2-acetamidoethyl)A2-thiazoline-4-carboxylate(3a), mp 156-1 58 OC, in yields up to 77% (purification by crystallization from benzenechloroform-petroleum ether or distillation at 160 OC/(O.l Journal of the American Chemical Society
a
(R’ = (C6H5)3C; 96%) as a white foam. Treatment with A g N 0 3 (1.3 equiv, pyridine-methanol, 12 h) at room temperature gave the corresponding silver mercaptide (10096, R’ = Ag) as pale yellow crystals. The mercaptide was converted to mercaptan 5b (loo%, R’ = H ) by treatment of a methanolic suspension of the silver salt with H2S: N M R (CDC13, ( C H 3 ) a i ) 6 1.33 (t, 3), 1.47 (t, l), 3.12 (dd, 2), 3.35 (t, 2), 3.88 (m, 2),4.27 (4,2), 4.98 (m, I ) , 7.3-7.5 (m, 3), 7.7-8.2 (m, 3), 8.50 (t, I), 8.96 (d, 1). Compound 5b (R’ = H ) was dissolved in CHC13 and treated with a slow stream of hydrogen chloride (36 h, room temperature). After concentration of the reaction mixture, the residue was partitioned between ethyl acetate and aqueous Na2CO3. Workup of the organic phase afforded a clear oil (90% recovery; A, (1: 1 C2H50H-HCl) 233 and 300 nm; presumably the thiazolylthiazoline) which was redissolved in CHC13 and shaken in the presence of Mn02 or NiO2’O (5 days, room temperature). Workup gave a yellow oil which deposited colorless needles of the known1’ ethyl 2’-(2-benzamidoethyl)-2,4’-bithiazole-4-carboxylate (6b) from ethyl acetate-petroleum ether: yield 24%; mp 143-144 OC; Amax (EtOH) 290 nm (log 4.17; NMR CDCIl, (CH3)dSi) 6 1.46 (t, 3), 3.36 (t, 2), 3.93 (t, 2), 4.47 (q, 2), 7.35-7.9 (m, 6), 8.06 (s, l), 8.19 (s, 1). Having obtained the desired bithiazole (6) via stepwise dehydrative cyclization and oxidation, it was of interest to attempt the direct conversion of @-alanylcysteinylcysteine derivative 7 to 6 via bithiazoline 8. Treatment of an ethanolfree CHC13 solution of 7a (R’ = H)12 with a slow stream of HCl(24 h, room temperature, followed by concentration under diminished pressure) afforded a water-sensitive residue having the UV spectrum (A, (1:l C2H50H-HCl) 266 nm (e 9200)) expected of bithiazoline 8a.13Attempted oxidation of the putative bithiazoline to 6a (Ni02, CHC13) gave instead the disulfide derived from 5a (R’ = H), whose formation may pro-
/ 99:24 / November 23, 1977
8079
ceed via hydrolysis of 8a by water associated with the oxidant or formed during the oxidation. Although NiOz could not be employed for the conversion 8a 6a, this reagent has also been used for the attempted oxidation of other partially reduced N-, 0-,and S-containing heterocycles, many of which were dehydrogenated in good yield. Compounds oxidized successfully with NiOz included 2-methylthio-A2-thiazoline (60%), methyl 2-methyl-A2-imidazoline-4-carboxylate (8 l%), 1,5-diphenyl-3-(p-bromopheny1)pyrazoline (95%),14 2,3-dihydrobenzofuran (52%), and several 2,4-disubstituted A2-thiazolines, including phleomycin A2 (83%).16
-.
Acknowledgments. W e thank Professor George Buchi and Dr. John Vederas for helpful discussions during the course of this work, Dr. Robert Engle (National Cancer Institute) for providing an authentic sample of 2’-(2-aminoethyl)-2,4’bithiazole-4-carboxylic acid, and Dr. H. Umezawa for samples of phleomycin A2 and bleomycin A2. This investigation was supported by contract N01-CM-437 12 from the Division of Cancer Treatment, National Cancer Institute, National Institutes of Health, Department of Health, Education, and Welfare. References and Notes (a)H. Umezawa, Rug. Biochem. Pharmacoi., 11, 18 (1976); (b)T. Ichikawa. ibid., 11, 143(1978):(c)S. K.CarterandR.H.Blum, ibid., 11, 158(1976); (d) G. Bonadonna, G. Tancini, and E. Bajetta, ibid., 11, 172 (1976); (e) A. Depierre, ibid., 11, 195 (1976); (f) J. Rygard and H. S . Hansen, ibid., 11, 205 (1976); (9) P. Rathert and W. Lutzeyer, ibid., 11, 223 (1976). See G. E. Hail, N. Sheppard, and J. Walker, J. Chem. SOC.C, 1371 (1966), and references therein. The peptide antibiotic bacitracin A, e.g.. has a cysteinyl residue that exists preferentially as the corresponding thiazoline (E. P. Abraham and G. G. F. Newton, Biochem. J., 53, 604 (1953); L. C. Craig, W. Hausmann, and J. R. Weisiger, J. Am. Chem. SOC.,78, 2839 (1954)) and the production of micrococcin P ceased when cysteine was omitted from the fermentation medium (P. Brwkes, R. J. Clark, A. T. Fuller, M. P. V. Mijovic, and J. Walker, J. Chem. Soc.,916 (1960)). Further support for this suggestion may be inferred from bleomycin and phleomycin, which are both produced by Streptomyces verticiilus and differ structurally only in a single double bond, such that phleomycin (which contains a thiazoiinylthiazole) may be regarded as the biosynthetic precursor of bleomycin. Also, the substitutents and substitution patterns associated with all of the cited compounds are consistent with their derivation from peptides. E.g., (a) I. Goodman and L. Sake, Biochim. Biophys. Acta, 100, 283 (1955); (b) Y. Hirotsu, T. Shiba, and T. Kaneko, ibid., 222, 540 (1970). See. however, Y. Hirotsu, T. Shiba, and T. Kaneko, Bull. Chem. SOC.Jpn., 40, 2950 (1967). E.g., (a) bottromycin, J. M. Waisvisz, M. G. Van der Hoeven, and B. te Nijenhuis, J. Am. Chem. Soc., 79, 4524 (1957); (b)althiomycin, B. W. Bycroft and R . Pinchin, J. Chem. SOC., Chem. Commun., 121 (1975), and references therein; (c) siomycin, M. Ebato, K. Miyazaki, and H. Otsuka, J. Antibiot. (Tokyo), 22,423 (1969), and Y. Wikasaka, T. Nagasaki, and H. Minato, ibid., 26, 104 (1973); (d) thiostrepton. B. Anderson, D. C. Hodgkin, and A. Viswamitra, Nature, 225, 233 (1970); (e) zorbamycin, Y. Ohashi, H. Abe. S.Kawabe, and Y. No, Agr. Bioi. Chem., 37, 2387 (1973); (f) saramycetin, A. Aszalos, A. I. Cohen, J. Aiicino, and B. T. Keeler, Antimicrob. Agents Chemother., 456 (1967);(9)nosiheptide, H. Depaire, J.-P. Thomas, and A. Brun, Tetrahedron Lett., 1403 (1977). See also J. Jadot, J. Casimir. and R. Warin, Bull. SOC.Chim. Beig., 78, 299 (1969). E.g., (a) phosphorus pentachloride. S. Gabriel, Chem. Ber., 24, 1110 (1891), and S.Gabriel, ibid., 49, 1110 (1916); (b) phosphorous oxychloride, J. C. Vederas, Ph.D. Thesis, Massachusetts Institute of Technology, 1973. Compound 2a (R‘ = H) was obtained (100%) by successive treatments of N-acetyl-b-alanyl-S-tritylcysteine ethyl ester (2a, R‘ = (C6H&C) with HgClp or AgN03 and then with methanolic H2S. The fuiiy blocked dipeptide was prepared by condensation of Kacetyl-&alanine and S-trityicysteine ethyl ester (79 % ; DCC, N-hydroxysuccinimide). E.g., (a) potassium ferricyanide and mercuric acetate, J. Walker, J. Chem. SOC.C, 1522 (1968), and N. A. Fuller and J. Walker, ibid., 1526 (1968); (b) hydrogen peroxide and potassium dichromate, F. Asinger, M. Thiel, and L. Schroeder. Justus Liebigs Ann. Chem., 810, 49 (1957); (c) cupric SUIfate;5b(d) MnOp and various quinones, M. A. Barton, G. W. Kenner, and R. C. Sheppard, J. Chem. SOC.C, 1061 (1966). Containing 2.83 mequiv of active 0 2 , as measured by iodide titration: K. Nukagawa. R. Konaka, and T. Nakata, J. Ofg. Chem., 27, 1597 (1962). New compounds gave satisfactory elemental analyses or high resolution mass spectra. The two oxidants gave comparable yields for this transformation. (C2H50H)290 nm (log t 4.18), has Compound 8b, mp 143-145 OC, A,, been prepared previously: K. Y. Zee-Cheng and C. C. Cheng, J. Heterocyci. Chem., 7, 1439 (1970). Obtained from 2a (R’ = (CgH&&) in 67% overall yield via saponification (methanolic NaOH, reflux, 10 min), condensation with S-tritylcysteine ethyl ester (DCC, Khydroxysuccimmide) to give 7a (R’ = (C6H5)3C)and deblocking via the monomercuric mercaptide.
Alternate workup procedures, which permitted the product to come in contact with an aqueous phase, gave material with much smaller apparent abswptivitiy. Although c is dependent on the nature of the ring substituents and the medium in which the UV spectrum is recorded, molar absorptivity values of 5000 for single thiazolines are typical. See, e.g., (a) W. Stoffel and L. C. Craig, J. Am. Chem. Soc., 83, 145 (1961); (b) W. Konigsberg and L. C. Craig, J. Am. Chem. SOC., 81, 3452 (1959), and ref 7a. K. S.Balachandran, I.Bhatnagar, and M. V. George, J. Org. Chem., 33,3891 (1968). This oxidation was carried out by Mr. David Evans. Conversions of phleomycin D, and E to bleomycins B2and B4,respectively, have been reported previously in unspecified yields. See T. Takita, Y. Muraoka. A. Fujii, H. Itoh, K. Maeda, and H. Umezawa, J. Antibiot. (Tokyo), 25, 197 (1972); H. Umezawa, Biomedicine, 18, 459 (1973). (17) Fulbriaht-Havs Scholar. 1975-1 976. (18) National Cancer Institute Postdoctoral Trainee, 1975-1977. (19) National Cancer Institute Career Development Awardee, 1975-1980. Alfred P. Sloan Research Fellow, 1975-1979. John Simon Guggenheim Fellow, 1977- 1978.
Donald A. McCowan, Ulrich Jordis” David K. Minster,’* Sidney M. Hecht*I9 Department of Chemistry Massachusetts Institute of Technology Cambridge, Massachusetts 02139 Received June 6, 1977
Oxidation of 9-Hydroxy- and 9-Methoxyfluorene Carbanions by Flavin. Proof of Radical Mechanism Sir: Flavin mediated dehydrogenation reactions which introduce unsaturation a,/3to carbonyl groups are of considerable biochemical interest (lactic acid oxidase, amino acid oxidases, succinic acid dehydrogenase, etc.) and have been the subject of numerous Model studies from this laboratory2b,3b,dhave firmly established that it is the resonance stabilized carbanion of the substrate which undergoes oxidation by flavin. Kinetic and other evidence supports a radical mechanism (Scheme IA) or, less likely, a mechanism involving a 4a adduct which goes on to product by specific base catalysis (Scheme IB).zb,3,4 The mechanism of Scheme IA has been favored3on the basis of free-energy c a l c ~ l a t i o n sarguments , ~ ~ ~ ~ centered around the requirement of specific base catalysis of 4a-adduct decomposition,3d and the results of studies with 1,5-dihydro-3,5dimethyll~miflavin.~ However, direct evidence for the formation of a flavin-substrate radical pair, as required by Scheme IA, has not been obtained. The present study deals Scheme I OH
I
k , [El
I
% [BHI
R-C-X
H
YH
,
R-C-X
r-)
rds
k3
a ’ FLH-
+
OH
8
I
0-
R-C-X
k-3
OH
B)
R-h-X
+ Flo,
(-)
8
+ R-C-X
Communications to the Editor
